Projects
The increasing usage of social media platforms for social expression leads to the generation of large amounts of unstructured text in the form of messages, posts, and blogs involving politics and political discourse. In addition to the expression and exchange of ideas and information, social media is a remarkable platform to get validated for your ideas as well gain popularity when liked by a large set of users. One such platform is Twitter which generates huge amounts of text containing political insights, which can be mined to analyze the kinds of tweets that are more likely to be from Democrats vs from Republicans thus allowing us to understand trends in the thoughts of Democrats vs Republicans on social media. In this work, an attempt is made to analyze collected twitter data and predict the political party associated to the text by modeling the problem as a supervised learning problem, namely, a classification problem.
The increasing discourse about climate change in the recent years has been leading the world to focus on a major energy transition. This increase in global ambition has shifted the pace of change in the energy system and has been impacting the energy mix. However, the world has historically been (and still is) heavily reliant on the use of fossil fuels and non-renewable sources. In this paper, we explore the consumption of renewable and non-renewable energy sources since 1965-2019 as well as the energy mix of each type - hydroelectric, solar, and wind for renewable energy and oil, gas, and coal for non-renewable energy. We also apply time-series analysis algorithms on the US renewable and non-renewable energy consumption and compare the predicted growth to that of the rest of the world.
Studying exoplanet atmospheres give us insight into the composition of exoplanets as well as habitability. Existing multi-object spectrographs are commonly used for medium-resolution transit spectroscopy observations; however, these instruments are not optimally efficient and often are affected by instrument systematics that are difficult to calibrate out of the data. The High-efficiency Instrument for Rapid Assessment of eXo-atmosphere (HIRAX) is an upcoming instrument for Palomar that can image a star in multiple high throughput, narrow bandpasses achieving R~\(2000\), and is being specifically designed for atmospheric characterization of exoplanets. In this paper, we describe work simulating HIRAX observations to understand the optimal choice of the instrument’s five bandpasses, including their center wavelengths and bandwidths in order to detect sodium in transiting Hot Jupiter atmospheres. We present the results of signal-to-noise ratio calculations describing the expected sodium detection achievable for host stars of varying magnitudes for the various bandpass configurations as well as estimations of the number of confirmed exoplanets that could be characterized for each configuration. Finally, we will show the effects of Doppler shifts on the exoplanet spectrum with respect to the static HIRAX bands due to various system velocities and how the band placements overlap with characteristic features of the Earth’s atmosphere.
In this project we observe the transit of the exoplanet Qatar-1b using the 24-inch telescope situated in the MEES observatory. With our observational data of the planet we determine its orbital period, \(T = 26.938/n\) days with n being the number of orbital cycles apart we observe the planet (from comparing with exoplanet database we find n = 19), and radius, \(R = 1.29\pm 0.05 R_{\textrm{Jupiter}}\) while considering the effects of limb darkening on the light curve. Additionally we estimate its inclination angle to be around \(90^{\circ}\) as we could observe its transit. Using its physical properties we also inferred that it is a hot Jupiter that is too close to its host star making its surface extremely hot and uninhabitable. Given its size and orbital radius we also inferred that it could not have formed so close to its star and must have drifted towards it while accreting matter. We finally compared our calculated planetary properties with those found in databases and found our T and R are within \(3\sigma\) of the database values of \(P = 1.4\) days and \(R_{0} = 1.14 R_{\textrm{Jupiter}}\).